Reading the Trades: Exploring Umia’s first Decision Market
Umia put its $4.77M treasury to a decision market. What 51 traders did across four outcomes, and what their trades record that a vote never could.
Between September 17 and 19, Umia put its $4.77M treasury to a decision market. There were four outcomes: keep all USDC idle (the no-op market), or deploy it into lending on Base under one of three allocation strategies. Each outcome traded as its own conditional UMIA market for 48 hours. The winner would be the outcome with the highest time-weighted average price (TWAP) over the window, provided it beat the no-op by at least 2%. Full 50/50 on Aave & Steakhouse won with a margin of +6.75%. The market settled at 13:04 UTC on September 19 and the funds were deployed at 17:09 UTC, about four hours later.
53 wallets traded. Two were arbitrage bots, so the figures below cover the other 51, referred to as traders.
Every trade in a decision market is conditional and onchain, so the result comes with a record a vote can't produce: what each participant wanted under each outcome, and how much they committed to it.
The market ranked every option
| Outcome | Settlement TWAP (USDC per UMIA) | Margin over no-op |
|---|---|---|
| Full 50/50 on Aave & Steakhouse (won) | $0.6704 | +6.75% |
| Half deployment: Aave & Steakhouse | $0.6481 | +3.21% |
| Full: Aave & Gauntlet + Steakhouse | $0.6310 | +0.49% |
| No-op (keep all USDC idle) | $0.6279 | Baseline |
Two of the three deployment options cleared the 2% threshold. Half deployment would have passed on its own; it lost only because Full 50/50 ranked higher.
Comparing pairs of outcomes shows what traders cared about. The two full deployments are the same size but use different curators, and their margins differ by a very significant 6.26 percentage points, pushing one below the passing threshold and another one above. Full 50/50 and Half deployment use the same venues at different sizes, and their margins differ by only 3.54 points. Traders put more weight on where the funds went than on how much of the treasury was deployed.
How traders split across outcomes
Most traders used the market much like a ballot, picking one option and backing it. 45 of the 51 traded a single outcome: 30 bought only Full 50/50, 4 bought only Full with Gauntlet, 3 bought only Half deployment, and 6 only sold the no-op. The remaining 2 bought only the no-op, making them the only traders who directly backed keeping the treasury idle. Just 6 wallets traded more than one outcome.
Buying also dominated. 41 of the 51 traders only ever bought, 8 only ever sold, and 2 did both. Of the 11 wallets that sold at all, 10 sold the no-op, so almost every seller was betting against leaving the funds idle.
A vote against the status quo, with no opinion on the rest
Selling UMIA in the no-op market is a bet that doing nothing is the worst option. 10 of the 51 traders sold in the no-op market. Six of them did nothing else: no buys, and no trades on any of the three allocations.
Leaving out the largest wallet (covered below), the other nine no-op sellers were all small wallets, and together they sold $13,760. These traders expressed no preference between allocations. They were pricing "anything but idle".
One deposit, two plans
The largest wallet deposited 30,000 UMIA and $5,157 USDC. It sold UMIA in three markets (no-op, Half deployment, and Full with Gauntlet) for $43.4K in total, and bought UMIA in Full 50/50 for $5.2K.
In practice this was a contingent plan: if Full 50/50 wins, buy about 7.5K more UMIA; if any other outcome wins, sell most of its 30K. Only trades in the winning outcome settle, so the sells were voided and the buys filled at an average of $0.682. The wallet ended with 37,563 UMIA.
Its final trade came two minutes before close and set the market high of $0.7065. Because the winner is decided on the average price across all 48 hours, a trade that late had little effect on the result.
Who Showed up to the Market?
Looking overall at the trader participation numbers, the participation numbers show a few patterns:
- Participation was concentrated. The largest wallet traded 64% of trader volume, and the 10 largest traded 96%. The median voter traded $91 in total.
- Most traders came once. 42 of the 51 made a single trade and never returned, and 30 of the 51 did nothing but buy the eventual winner.
- Activity was front-loaded. 62% of trader volume traded in the first 6 of the 48 hours. In the final two hours, voters traded $16.8K, of which $16.6K came from the largest wallet.
- Most traders were not bidders in the UMIA auction. Only 19 of the 51 had bid in the UMIA auction, which is 2.8% of its 678 bidders.
- Liquidity held. Each outcome was seeded with 1.30M UMIA and $800.8K, and the largest single swap moved its outcome's price by 1.56%.
- Backing a losing outcome cost nothing. Deposits in losing outcomes are returned 1:1, so the 9 of 51 traders who only backed losing options get their funds back in full.
When looking at it, the data contains three things no vote produces: a full ranking with margins, a rejection of the status quo recorded separately from the choice of allocation, and conditional plans executed from a single deposit. Cohort 1 ventures will run their structural decisions through the same markets and generate the same kind of record.
Disclaimer: This article is meant for informative purposes only. When trading in any market, returns are not guaranteed and losses may happen.